A clustered caching placement in heterogeneous small cell networks with user mobility

We consider the content caching problem in a set of the clustered small cells (SCs) in the heterogeneous network. Heterogeneous small cell networks is a promising solution to overcome the explosive growth of the mobile data traffic. A bottleneck of this solution is the capacity of backhaul links between small cells and the network core. However, an alternative solution is using SCs with low rate backhaul links which are equipped with storage devices to cache the popular files. We group SCs into some clusters based on the geographical situation. Each cluster has a cumulative cache storage which is the summation of the cache storage of SCs encountered at that cluster. During one time slot, users in a cluster can download the desired files from the cumulative cache storage of the cluster. Since users may move and change their positions, we model the problem of cluster cache placement as a discrete Markov chain. As the optimum caching placement problem is NP-hard, we provide an appropriate lower bound that guarantees k optimal solution where k is the random walk step of the user's movement. Further, we derive a polynomial time approximation solution with performance for a situation which the information about the mobility pattern of mobile users in the wireless cell does not exist. Through some simulation examples, we compare the proposed algorithms with an existing solution using real wireless data.